Introducing GAN Dissection (http://gandissect.csail.mit.edu/), a framework for visualizing and understanding generative adversarial networks. Our methods let you explore what a GAN has learned by examining and manipulating its internal neurons. Using our interactive tools, you can paint using neurons instead of pixels.
David Bau, Jun-Yan Zhu, Hendrik Strobelt, Bolei Zhou, Joshua B. Tenenbaum, William T. Freeman, Antonio Torralba. GAN Dissection: Visualizing and Understanding Generative Adversarial Networks. arXiv preprint arxiv 1811.10597, 2018.